ESTROGEN A N D CHICK OVIDUCT
DNA
RENATURATION
Effect of Estrogen on Gene Expression in the Chick Oviduct. I. Deoxyribonucleic Acid-Deoxyribonucleic Acid Renaturation Studiest Jeffrey M. Rosen, Charles D. Liarakos, and Bert W. O’Malley
As a prerequisite for studying the effects of estrogen on the transcription of unique sequence DNA during hormone-mediated growth and differentiation of the chick oviduct, it was necessary to first study the complexity of chick DNA and secondly fractionate the chick genome into repeated and unique sequences. Approximately 30% of the chick genome consists of rapidly renaturing or repeated sequences, with the remainder being the more slowly renaturing, unique sequences. No effect of hormone treatment on the kinetics of chick oviduct renaturation was experimentally detectable, suggesting that estrogen-induced differentiation was not accompanied by major gene amplification or deletion. Identical renaturation profiles of 3H-labeled chick embryo and chick liver DNA also indicate that tissue differentiation during embryogenesis did not give rise to major gene amplification or deletion. Tritium-labeled unique sequence DNA of high speABSTRACT:
A
dministration of diethylstilbesterol or estradiol to a newborn chick results in the differentiation of the immature oviduct’s mucosal cells into three distinct epithelial cell types, two of which synthesize specific marker proteins (Kohler et al., 1968; O’Malley et al., 1969; Oka and Schimke, 1969). Previous studies from this laboratory have suggested that the estrogen-mediated growth and differentiation of the chick oviduct involve an alteration in gene expression, leading to the induction of new species of hybridizable nuclear RNA (O’Malley and McGuire, 1968.) However, hybridization of nucleic acids from higher organisms, performed under the standard conditions of low nucleic acid concentrations and short incubation times, has been shown to involve only the rapidly reassociating repetitive sequences to the exclusion of the more slowly reassociating sequences, operationally defined as unique sequences (Britten and Kohne, 1968). Furthermore, the presence of closely related sequences in eukaryotic DNA may result in a lack of gene locus specificity, i.e., mismatching of base sequences (McCarthy and Church, 1970). Recent advances in the technology of nucleic hybridization from higher organisms have permitted the study of transcription of unique sequence DNA in eukaryotes (Gelderman et al., 1971 ; Melli et al., 1971; Grouse et al., 1972). In order to investigate the effects of estrogen on the transcription of unique sequence DNA, it has been necessary to first study the kinetics of chick DNA renaturation, i.e., chick DNA complexity, and secondly subdivide the chick oviduct genome on
t From the Department of Cell Biology, Baylor College of Medicine, Houston, Texas 77025. Receiced February 21, 1973. This work was supported by National Institutes of Health Research Grant HD-04473, American Cancer Society Grant BC-lOlB, and Ford Foundation Grant 630-014 1,
cific activity was isolated from chick fibroblasts and analyzed by renaturation with an excess of unlabeled chick DNA. Complete sequence homology existed between the unique sequences of chick fibroblast DNA and the unique sequences in chick liver DNA. Unique sequence DNA was also labeled chemically by an ultraviolet-catalyzed photoreduction with tritiated NaBH4. Only 40% of the tritium radioactivity in this DNA renatured to form a stable duplex with an excess of unlabeled chick DNA, suggesting an alteration in the physical properties of the chemically labeled DNA. However, melting profiles on hydroxylapatite indicated little mismatching and the formation of stable duplexes between chick DNA and both chemically labeled and chick fibroblast DNAs. These results allow for the investigation of unique sequence transcription during estrogen-mediated changes in the chick oviduct.
the basis of nucleotide sequence homology into unique and repeated sequences. An attempt was also made to determine whether estrogen treatment results in major gene amplification or deletion during hormone-mediated growth and differentiation of the oviduct. Selective amplification of the repetitive DNA in several eukaryotes has recently been reported (Gall, 1969; Willie, 1972). In addition, as a prerequisite for performing saturation hybridizations under conditions of vast RNA excess (Grouse et al., 1972), it was necessary to prepare and characterize labeled unique sequence DNA of high specific activity. The results of DNA-DNA renaturation experiments using tritium-labeled DNA prepared by several different labeling methods are presented. Analysis of the kinetic complexity of various RNA fractions and saturation hybridizations with RNA populations extracted from the oviduct at various stages of differentiation will be reported in the subsequent papers in this series. Experimental Procedures
DNA Preparation and Labeling. All DNAs were extracted from nuclei by a modification of the method of Marmur (1961) which consists of the addition of a polysaccharide hydrolysis step using a-amylase (Worthington) at a concentration of 50 pg/ml for 30 min (37”) and the omission of the isopropyl alcohol precipitation procedure. Routine analysis of DNA involved an optical density scan and thermal denaturation profile determination (Mandel and Marmur, 1968) using a Gilford 2400 spectrophotometer and a diphenylamine assay (Burton, 1968). Escherichia coli B late-log and Bacillus subtilis SB 19 were purchased as frozen cell pastes from Miles Laboratories. Tritium-labeled B . subtilis DNA was prepared from late-log cells grown in Spizizen’s Minimal Salts (Sober, 1968) supplemented with 10% casein hydrolysate and 10% B I O C H E M I S T R YV O, L . 1 2 ,
NO.
15, 1 9 7 3
2803
ROSEN, LIARAKOS, A N D O ' M A L L E y
7
I
2o
t
1
10.)
1
IO"
IO"
100
IO'
IO'
103
IO'
Cot (mole s e c / I i t c r )
FIGURE 1 : Renaturation kinetics of 20-day 0.14 M phosphate at 62" for oviduct DNA
estrogen-treated chick oviduct DNA and 5.subfilis DNA. Renaturation experiments were run in (0)and 3H-labeled B. subtilis DNA (A). Conditions were as described under Experimental Procedures. Coti/2values were: B. subtilis DNA, 1.5; chick oviduct DNA, 300. glucose, and labeled with 5 pCi/ml of 5-[methyL3H]thyrnidine (20 Ci/mmol, New England Nuclear} during log phase. E . coli DNA was extracted directly from the saline-rinsed cells. 3H-Labeled chick fibroblast DNA was isolated from a primary monolayer culture of chick fibroblasts purchased from Microbiological Associates, and labeled from 30% monolayer to confluency with 20 pCi:'ml of 6-E3H]thymidine (9.1 Ci/mmol, SwartzjMann). For the preparation of 3H-labeled chick embryo DNA 125 pCi of 6[ 3HH]thymidine (9.1 Ci/mmol, SchwarziMann) was injected into the air sac of Marek negative specific pathogen-free fertile eggs (Spafas, Inc.) on days 7 and 8 after incubation at 37" was begun and the DNA was extracted on day 9. Chemical labeling of DNA with tritiated NaBH4 was performed according to the procedure of Kirkegaard (1969), as modified by Vincent et ul. (1969). Unique DNA (sheared and denatured), isolated as described in a following section, was dissolved in 0.01 M Tris-IICI, pH 8.5, at a concentration of 1.5 mg/ml. Lyophilized tritiated NaBH4 (15 Ci/mmol, New England Nuclear) was then added to yield a final concentration of 100 mCi/ml and the solution irradiated at a 1-cm path length with mineralight UVS-I1 (Ultraviolet Products) in a 3-ml quartz cuvet with constant stirring for 6 min. The reaction was stopped by the addition of 6 drops of butanone and the solution stirred for an additional 3 min. All these procedures were carried out in a fume hood. Following three extractions with an equal volume of ether, the DNA was precipitated from ethanol several times until constant specific activity was attained. The DNA was further characterized by its binding to hydroxylapatite (Clarkson) and elution with 0.14 M sodium phosphate buffer, pH 6.8 (prepared by mixing equimolar amounts of Na2HP04and NaH?P04).Treatment with RNasefree DNase I(50 pg/ml at 37" for 1 hour) rendered 9 4 x of the radioactivity ClaCCOOH soluble, indicating that the tritium was incorporated primarily into DNA and not into a small amount of contaminating protein or polysaccharide. Renaturation Kinetics of DNA. DNA renaturation was measured on hydroxylapatite columns by the method of Laird (1971), as adapted from the original procedure of Britten and Kohne (1968). DNA was sheared in 0.015 M NaC1-0.0015 M sodium citrate, pH 7.0 (0.1 x SSC), at 16,000 psi in a French pressure cell press (Aminco} to yield fragments approximately 300-400 nucleotides in length (single stranded). The 0.1 X SSC solution was then made 0.15 M in Na+ by the addition of NaCl and the DNA alcohol precipitated. The resulting DNA pellet was dissolved in 0.14 M phosphate buffer to a concentration of between 3 and 8 mg/ml and dialyzed extensively
2804
BIOCHEhlISTRY,
VOL.
12,
NO.
15, 1 9 7 3
against 0.14 M phosphate buffer. In some experiments 3Hlabeled B. subtilis DNA (56,800 cpm/pg) was added at a concentration of 10-20 pg/ml (less than 1 % of unlabeled DNA) as an internal control (Laird, 1971), and the final DNA solution contained 0.001 M N-tris(hydroxymethy1)methyl-2aminoethanesulfonic acid buffer, pH 7.0. In some cases a dilute renaturation curve was also run at approxima:ely 120 pg/ml to determine the early section of the renaturation curve (Cot values
